The investigators take advantage of naturally occurring disease in the soft-shelled clam Mya arenaria. This leukemia: 1) is fatal; 2) occurs at high frequency in the wild (five-to-sixty percent); 3) is transferable to normal clams and; 4) is remarkably similar to human ALL-L3 (Burkitt's leukemia) in the morphology, behavior and cytochemistry of the leukemia cells, in the structure and expression of the p53 gene and in its response to etoposide. In preparation for this study, we have developed a mass culture system for clam leukemia cells and are the first to clone and determine expression patterns for clam c-ras and p53. Clam p53 has a binding site for MDM2 and its transcriptional activation domain is seventy-three percent conserved compared to human p53 protein. This suggests that downstream genes involved in the cell cycle and in apoptosis should also be transcriptional targets for clam p53. Much remains to be learned about this emerging leukemia model. We will use clam blood to address three specific aims that should expedite our understanding of genes involved in this disease and should help to either further validate or to discount this model. The investigators will: 1) Determine the cytotoxicity for clam leukemia of compounds widely employed in human ALL chemotherapy; 2) Determine the cytotoxicity of novel compounds that target p53 or related molecules involved in the cell cycle and/or apoptosis and; 3) Perform mutational analysis of clam p53, c-ras and c-myc. For the first two of these objectives, high throughput cytotoxicity screening will initially be accomplished using microtiter plates. Compounds that show activity against clam leukemia cells will be further tested using larger scale, parallel in vitro and in vivo clam assays for cell and organism viability, p53 expression and apoptosis. Mutational analysis using mismatch detection should reveal mutations in three genes that often cooperate during transformation inhuman leukemia. Clam leukemia offers significant advantages over currently available models for the following reasons: a) It provides an in vitro and in vivo alternative to the relatively few human leukemia cells lines; b) Populations of leukemic clams are more similar to an outbreeding, human clinical population than are those generated from inbred mouse strains or by intentional exposure to know tumor viruses; c) While existing fly and worm models have versions of some human cancer genes that contain naturally occurring or inducible mutations, resulting tumors characteristically affect embryonic and not adult somatic cells (which do not divide in these organisms) and; d) Clams have highly conserved homologs for human c-ras and p53 genes. Homologs for p53 have not yet been identified in any other non-vertebrate models, including yeast. Data that we generate in this study should point out p53-related molecular mechanisms that are held in common with human ALL.